Characteristics of the Vincristine-induced Augmentation of Methotrexate Uptake in Ehrlich Ascites Tumor Cells*

SUMMARY Studies were undertaken to characterize the effect of vincristine on methotrexate uptake in Ehrlich ascites tumor cells and to define the mechanism of this interaction. Vincristine (10 FM) does not alter the unidirectional influx of methotrexate but slows the unidirectional efflux of methotrexate and leads to a large increase in the steady state level of exchangeable intracellular methotrexate. These effects of vincristine are rapidly, but not completely, reversible. Vincristine stimulation of net methotrexate uptake occurs within 4 of exposure of cells to this and increases an extracellular vincristine concentration range of at least 5 to 50 FM. not the intracellular content nor the chloride distribution The addition of the the effect of vincristine and completely reverses the stimulatory effect of on the net uptake of methotrexate.

to characterize the effect of vincristine on methotrexate uptake in Ehrlich ascites tumor cells and to define the mechanism of this interaction. Vincristine (10 FM) does not alter the unidirectional influx of methotrexate but slows the unidirectional efflux of methotrexate and leads to a large increase in the steady state level of exchangeable intracellular methotrexate. These effects of vincristine are rapidly, but not completely, reversible. Vincristine stimulation of net methotrexate uptake occurs within 4 min of exposure of cells to this agent and increases over an extracellular vincristine concentration range of at least 5 to 50 FM. Vincristine does not alter the intracellular water content nor the chloride distribution ratio.
The addition of glucose to the medium partially reverses the effect of vincristine and completely reverses the stimulatory effect of sodium azide on the net uptake of methotrexate.
Vincristine (10 pM) increases net uptake of methotrexate in L1210 leukemia cells but does not have a significant effect on net methotrexate uptake in the L-cell mouse fibroblast over a 2f$-hour interval of exposure even though replication of the L-cell is completely arrested.
The similarity between the effects of vincristine and metabolic poisons on methotrexate uptake and the partial reversal of the effect of vincristine by glucose suggest that this agent enhances methotrexate uptake by the inhibition of cellular energy metabolism with the consequent inhibition of an energy-dependent process which limits methotrexate accumulation within the cell. The data suggest further that this effect of vincristine represents an interaction with a cellular element(s) that is different from the interaction which results in the arrest of cell division in metaphase. rabbit reticulocyte (4), and the L-cell mouse fibroblast.2 Unique to this transport system inhibitors of anaerobic and aerobic metabolism enhance rather than inhibit the net uptake of methotrexate into these cells (5). In L1210 leukemia cells sodium azide produced a small increase in the unidirectional influx of methotrexate, a large fall in the unidirectional efflux of exchangeable intracellular methotrexate with a resultant increase in the apparent intracellular electrochemical potential for methotrexate to a level considerably greater than that of the extracellular compartment.
The data suggested that metabolic poisons inhibit an energy-dependent process which drives methotrexate out of the cell (5) and a more extensive model was subsequently proposed to account for the over-all energetics of methotrexate transport (2). Recently, another laboratory reported that vincristine increases net uptake of methotrcxate into the L1210 leukemia cell (6). The present study characterizes the mechanism by which vincristine augments methotrexate uptake in the Ehrlich ascites tumor cell and suggests that vincristine acts as a metabolic poison to inhibit an energy-dependent process which limits the intracellular accumulation of methotrexatc. These findings relate to other st,udies from this laboratory which suggest that the augmented net uptake of methotrexate induced by vincristinc is associated with an increase in the inhibition of DNA synthesis by methotrcxatc (7).

EXPERIMENTAL PROCEDURES
Cells and Media-Cells employed in thcsc studies were: (a) the Ehrlich ascites tumor grown in CF1 mice; (b) Earles L-cell mouse fibroblasts maintained in cell culture in Y.L.E.3 medium supplemented with 5'$$ calf serum; and (c) L1210 leukemia cells grown in R.P.M.I. 1630 medium (8) supplemented with 5% fetal calf strum (sera obtained from Grand Island Xologicals, Grand Island, K.Y.).
Crlls in culture were grown in spinner flasks and were washed twiczc with 0" buffer prior to cspcrirncntation. Ehrlich ascites tumor cells were passed by weekly intraperitoneal inoculation of 0.2 ml of undiluted ascitic fluid. Mice were killed 6 to 12 days after inoculation, and the ascitic fluid was suspended into 0" buffer. Contaminating erythrocytes were removed in the supernatant fluid after two to three washes with 0" buffer and separation of the tumor cell pellet was achieved by centrifugation for 1 min at 250 x g.
The pH was maintained at 7.3 to 7.4 during incubations by passing warmed and humidified 95% 02-5y0 COs over the cell suspensions.
Cytocrits were under 3% and the temperature was 37".
Determination of Methotrexate Transport Kinetics and Intracellular Methotrexate Level-The unidirectional influx, efflux, and net uptake of [3H]methotrexate were measured as previously described (1,5).
Pertinent aspects of the experimental protocols are indicated in the legends.
Flux reactions were halted by injection of the cell suspension into 10 volumes of 0" buffered 0.85% NaCl solution (pH 7.4) and the intracellular methotrexate level was determined as follows.
The cell fraction was separated by centrifugation (2000 x g for 30 to 60 s) and washed twice with the 0" saline solution.
The washed cell pellet was aspirated into t)he tip of a Pasteur pipette, extruded onto a polyethylene tare, and dried overnight at 60". The dried cells were then peeled off the polyethylene tare. Pellets under 1 mg were weighed on the Cahn G-2 electrobalance (Cahn Instruments, Paramount, Calif.) and the other pellets which weighed in the range of 1 to 5 mg were measured on the Beckman LRI 800 automatic microbalance (Beckman Instruments, Fullerton, Calif.). The dried pellets m-ere placed in scintillation vials which were tilted at 45" in specially constructed racks so that the pellet would be totally encompassed by the subsequent addition of 0.1 ml of 1 N KOH. The cells were digested for 1 hour at 60" and cooled, 10 ml of scintillation mixture (I) was added, and radioactivity was determined with a Beckman LS-133 liquid scintillation spectrometer. 911 quench corrections were made employing [%-or [14C]toluene internal standards.
Determination of Intracellular Water and Chloride Distribution Ratio-Intracellular water was determined from the difference between the wet weight and dry weight of a cell pellet less t.he [W]inulin space. This procedure as well as the determination of the chloride distribution ratio (the ratio of the concentration of chloride in the intracellular water to its concentration in the extracellular water) has been described in detail (1,4 Division, Pearl River, I\;.Y.) mere purified by DEAE-cellulose-ion exchange column chromatography.
The column was eluted with a linear gradient of 0.1 to 3.4 M ammonium bicarbonate at pH 8.3 with subsequent lyophilization of the methotrcxate-containing fractions.
For the ident.ification of intracellular radiolabel, cells were incubated with tritiated methotrexate, then disrupted by sonic oscillation, and after a 9,000 x g centrifugat.ion for 20 min at 4", the supernatant was fractionated on a DEAE-cellulose column. These procedures have been published in detail (1).
Chemicals-Vincristine and vinblastine sulfate were supplied through the generosity of Dr. Robert' Hosley of the Eli Lilly Rcsearch Laboratories, Indianapolis, Ind. Sodium azide was ob-tained from Fisher Chemicals.

EJ'ect of Vincristine on Unidirectional
InJlux and Time Course of Methotrexate Uptake-Exposure of Ehrlich ascitcs tumor cells to 10 pM vincristine a few seconds prior to the addition of methotrexate results in a large increase in the steady state intracellular methotrexate level but the unidirectional influx of methotrexate is unchanged (Fig. 1). To assess the duration of the interval between addition of vincristine and the onset of stimulation of net methot.rexate uptake, cells were first incubated with methotrexate until a steady state was achieved following which a portion of the cell suspension was e,xposed to vincristine ( 1. The effect of vincristine (VCR) on the time course of methotrexate (MTX) uptake. Cells were exposed to 10 NM vincristine a few seconds before addition of methotrexate.
El.Or 2. The effect of vincristine (VCR) on net uptake of methotrexate (MY'X). Cells were brought to the steady state with 0.8 PM methotrexate.
At the arrow, a portion of the cell suspension was transferred to another flask containing vincristine to achieve a concentration of 10 PM.
x O*r  3), 30 PM vincristine did not alter the influx of methotrexatc while 10 mar sodium azide produced a 23% increase in methotrcxatc influx under the same conditions. Effect of Vim&tine on Unidirectional E#ux of Methotrexate-Cells were loaded with methotrexate to the steady state in the presence and absence of vincristine, then separated by centrifugation and resuspended into methotrexate-free buffer with or without vincristine, respectively, and the unidirectional cfflux of methotrexate was monitored (Fig. 4). Intracellular methotrexate in the control cells consists of two components.
There is an intracellular fraction (referred to as "exchangeable") that rapidly leaves the cell and another larger fraction that is bound within the cell. Vincristine may produce a small increase in a bound intracellular methotrexate component, as suggested from additional experiments in which the unidirectional efflux of methot,rexate was monitored over longer intervals.
The major effect of vincristine is a marked fall in the rate of exit of the eschangeable methotresate fraction with a 2%fold increase in the cfflux half-time.
Eflect of Vincristine on Intracellular Water, Chloride Distribution Ratio, and Estimated Electrochemical Potential for Intracellular J4ethotrexate-While 10 PM vincristine has a profound effect on methotrexate uptake, it has little effect on other physical properties of the cells under these experimental conditions (Table I). Vincristine does not alter the ratio of the intracellular water to dry weight or the ratio of the extracellular water to wet weight. The dry weight to wet weight ratio of a cell pellet is unchanged. Vincristine does not produce significant change in the chloride distribution ratio, suggesting that this agent does not alter the membrane potential under these experimental conditions. Table  II represents an analysis of the data of Fig. 4  A portion of the cell suspension was then exposed to 10 PM vincristine and lIptake in the control and vincristine-treated cells was continued for an additional 50 min until steady state conditions were achieved. The cell fractions were then separated by centrifugation and resuspended into a large volume of methotrexnte-free buffer. The vincristineexposed cells were again exposed to vincristine, the control cells were suspended in buffer alone. Further, while the distribution ratio for methotrexate in the absence of vincristine was only 29% more than that cx-  Cells were incubated with 1.0 PM methotrexate fbr 20 min following which aliquots of the cell suspension were transferred to other flasks containing vincristine.
Incubation was continued until the steady state for methotrexate was achieved following which four measurements of the intracellular methotrexate level were obtained.
The bars indicate the mean f S.E. of four replicate measurements from a representative experiment. The extracellular vincristine levels are indicated below the bars. The difference between control cells and cells exposed to 5 FM vincristine is significant to p < 0.02.
pected for a passive (nonenergy-dependent) process, the distribution ratio for methotrexate in the presence of vincristine exceeds that predicted for a passive process by a factor of 2.8.
Relationship between Vincristine Concentration and Augmentation of Net Transport of Methotrezate-In the experiment of Fig.  5, cells were brought to the steady state with 1 pM methotrexate following which aliquots of the cell suspension were transferred to other flasks containing vincristine and the intracellular metho- trexate level was measured when the steady state was again achieved. Stimulation of net methotrexate uptake was observed at the lowest vincristine level employed (5 pM) and increased through the highest vincristine level tested (50 PM).
Reversibility of Vincristine Eflect-The reversibility of the effect of vincristine on methotrexatc uptake was evaluated in terms of the alterations produced by this agent on the net uptake and the unidirectional efflux process. In the experiment of Fig. 6, cells were exposed to vincristine for 8 min and then were divided into two portions.
The cell fractions were separated by centrifugation; one portion was suspended into vincristine-free buffer while the other was resuspended into buffer containing this agent. The time course of methotrexate uptake was then monitored in these cells as well as control cells which had not been exposed to vincristine.
Stimulation of net methotrexate uptake by vincristine was markedly diminished after the agent was removed from the buffer, although a small increase in the net uptake of methotrexate did persist under these conditions. Fig. 7 illustrates the reversibility of the vincristine effect on the unidirectional efflux of methotrexate. Cells were incubated with methotrexate in the presence and absence of 10 pM vincristine for 20 min. The vincristine-exposed cells were divided into two portions and the cell fractions were separated by centrifugation and then resuspended in the presence or absence of 10 ELM vincristine.
Both, as well as control cells, were resuspended into methotrexate-free buffer. The unidirectional efflux of exchangeable methotrexate was increased in the cells resuspended into the vincristine-free buffer in comparison to the cells exposed to vincristine in the resuspension buffer but the efflux half-time remained higher, at least initially, than that of the control cells. The tightly bound intracellular methotrexate level reverted to that of the control upon exposure to the vincristine-free buffer. Effect of Glucose on Stimulation of Methotrexate Uptake by Vincristine or Azide-Glucose alone produces a small, but significant, (p < 0.05 from five experiments) depression of net methotrexate uptake (Fig. 8). Sodium azide (10 mM) markedly enhances the net uptake of methotrexate but this effect is completely reversed by glucose. Unlike azide, the stimulatory effect of vincristine on net methotrexate uptake is only partially reversed by glucose; however, it is clear that the inhibitory effect of glucose on net 6.0 Sons can be distinguished (p < 0.05 in five experiments). However, over a 2>h-hour exposure to 10 pM vincristine, changes in net methotrexatc uptake are negligible. Ten determinations of the cell methotrexate level were obtained after control and vincristine-exposed L-cells were at the steady state with methotrexate.
The averages of these values from five such experimcnts, done on 5 different days were, in turn, averaged.
Vincristine resulted in a 5% increase in the total cell methotrexate level, a value t)hat showed poor statistical significance (p > 0.1) ATION when the differences between the control and the vincristinetreated cells mere analyzed by a two-tailed t-test. At higher vincristine levels (30 PM), a significant increase in the cell methotrcxate level could be detected.
Vincristine at 10 PM produced a prominent stimulation of net methotrexate untake into the L1210 leukemia cell consistent with a report from another laboratory (6). Recovery oj Intracellular Radiolabel in Preseme of Vimristine-To exclude the possibility that vincristine-induced alterations in net uptake of methotrexate are related to chemical modifications of the methotrexate molcculc, DEAE-cellulose column chromatography was performed after sonic oscillation of cells exposed to methotrexate for 1 hour in the presence and absence of 10 PM vincristine.
In both cases more than 98% of the radiolabel added to the column was recovered in the methotrexate peak. methotrexate uptake in the presence of vincristine is greater than the inhibitory effect of glucose alone (p < 0.02 from five experiments) Eflect of Vinblastine on Net Methotrexate Uptake-Vinblastine, at a concentration of 10 pM, stimulated net methotrexate uptake but the magnitude or characteristics of this effect were not compared with that of vincristine.
Effect of Vincristine on Methotrexate Uptake and Cell Replication in L-Cell Mouse Pibroblasts and on Uptake of Methotrexate into LiHO Leukemia Cells-Followin g exposure of L-cells in their growth medium to 10 pM vincristine, cell replication ceases immediately and, within 2 hours, a significant difference between the cell count in the control and in the vincristine-treated suspen- Previous studies from this laboratory demonstrated that the net uptake of methotrexate into many mammalian cells is opposed by energy-dependent processes and that, when these processes are blocked with metabolic poisons, net uptake of methotrexate is enhanced (2,4, 5). This phenomenon has been related more specifically to the energetics of methotrexate transport with the proposal that metabolic poisons inhibit an energy-dependent process which drives methotrexate out of the cell (5). To account for continued uphill transport of methotrexate into the cell even in the presence of metabolic poisons, a model was presented which suggests that the asymmetrical distribution of organic phosphates across the cell membrane results in a countertransport in which the downhill flow of organic phosphates out of the cell via the methotrexate carrier system drives methotrexate uphill into the cell, a phenomenon which might be relatively unaffected by the immediate consequences of metabolic poisons (2). These studies were undertaken to clarify the mechanism by which the vinca a,lkaloids enhance net uptake of methotrexate into tumor cells (6) and to characterize the physical state of the intracellular methotrexate accumulat,ed in the presence of these agents. Upon exposure to vincristine, a small t)ightly bound intracellular methotrexate component may appear which is rapidly and completely reversed when vincristine is removed; this is under further study. The major effect of vincristine is an increase in the level of exchangeable intracellular methotrexate; this is to a large extent,, but not completely, reversed when vincristine is removed.
These changes occur without an alteration in the intracellular water content. Since vincrist,ine did not change the chloride distribution ratio, the data suggest that this agent did alter the membrane potential under these experimental conditions.
The chloride distribution ratio as measured in these experiments may not reflect the absolute membrane potential but is employed only to indicate relative changes in membrane potential (11,12). If the assumption is made that exchangeable methotrexate is osmotically active within the intracellular water, and not in part loosely bound (loose binding is, however, virtually impossible to exclude), then the data suggest that vincristine increased the intracellular electrochemical potential for methotrexate.
The effect of vincristine on the interaction between methotrexate and Ehrlich ascites tumor cells is comparable to that observed for azide, an inhibitor of aerobic metabolism, in these cells as well as in the L1210 leukemia cell (5). Both agents depress the unidirectional efflux of methotrexate, increase the exchangeable intracellular methotrexate fraction, and appear to increase uphill transport into the cell. However, while aaide produces a small increase in the unidirectional influx of methotrexate, such a change could not be demonstrated for vincristine at levels up to 30 ELM. Many structurally diverse compounds with different sites of action, iodoacetate, azide, antimycin A, dinitrophenol, inhibit energy metabolism and enhance net uptake of methotrexate (a similar effect was observed with anaerobiosis) (5). The metabolic effects of agents which inhibit aerobic metabolism should be antagonized by glucose as a result of glycolytic production of ATP which compensates for impaired oxidative ATP formation.
Accordingly, alterations in the sodium and potassium content of Ehrlich ascites tumor cells induced by antimycin A, and the reduction in the cellular ATP level induced by amytal or dinitrophenol, are eliminated by glucose (13). Likewise, glucose completely reversed the effect of azide on met,hotrexate uptake in these studies. Hence, it is of particular interest that glucose partially reversed the vincristine-induced increase in methotrexate uptake.
The similarity between the effects of azide and vincristine suggests that vincristine inhibits energy metabolism and that it is by this mechanism that this agent alters methotrexate uptake. Because glucose partially reverses the effect of vincristine, the data suggest that vincristine may inhibit aerobic energy metabolism; however, since glucose reversal is not complete, vincristine may also inhibit anaerobic energy metabolism.
Additional evidence that vincristine may inhibit energy-dependent processes comes from the observation that this agent like other metabolic poisons reduces the uphill transport of the nonmetabolized amino acid analogue cy-aminoisobutyric acid, into Ehrlich ascites tumor cells.4 Inhibition of both anaerobic and aerobic energy metabolism by vinca alkaloids, at concentrations comparable to those employed in these studies, has been reported for L1210 and Ehrlich ascites tumor cells (14).
Periwinkle alkaloids bind with high affinity (15) to protein subunits of microtubules to form protein-drug precipitates (16-18) which arc thought to be the filamentous structures and crystals that are observed on electron microscopy (19)(20)(21).
This interaction results in impaired formation of mitotic spindles (22-24) and in the arrest of cell division in metaphase (25, 26) which is considered to be the basis of the cytotosicity of these agents. It is unlikely that impaired spindle formation can account for the effects of vincristinc on methotrexate uptake. I-Icnce, although 10 PM vincristinc or vinblastinc produce crystal formation on electron microscopy within 55 hour after exposure of L-cell mouse fibroblasts t.o these agents (19), over this interval there are no significant changes in methotrexate uptake. Further, alt.hough the L1210 leukemia and the Ehrlich ascitcs tumors are very sensitive to the stimulatory effect of vincristine on methotrexate uptake, vincristine negligibly affects survival of mice bearing these tumors (6, 27) and over an interval of exposure of L-cells to vincristine sufhcicnt to arrest cell replication a significant change in methotrexate uptake is not demonstrable. Vinca alkaloids interact with other microtubular proteins. Marginal bundles of microtubular elements have been noted beneath the plasma membrane of some cells and are eliminated by exposure to vincristine (28). Microfilamentous elements procipitable by vinblastine have been extracted from erythrocyte membranes and exposure of erythrocytes to vinblastine produced an increase in sodium flux and glycolysis (29,30).
These structures may be involved in the regulation of cellular energetics in general and/or the energetics of membrane transport processes in particular. However, if vinca alkaloids bind to these microtubular elements with the same affinity that they bind to other microtubular proteins, this would not account for the changes in methotrexate uptake. Hence, although vincristine or vinblastine bind rapidly to porcine brain microtubular protein, the rate of dissociation of the complex is slow (half-time, 5 hours) (15), while the effects of vincristine on methotrexate uptake are more rapidly reversible. Although periwinkle alkaloids interact with other cellular elements, this usually requires levels of these agents which are orders of magnitude higher than concentrations which alter methotrexate uptake (31).
Vinca alkaloids produce other changes in membrane function. Vinblastine inhibits phagocytosis in rabbit polymorphonuclear leukocytes (32) and augments membrane internalization in erythrocytes (33). Although vinca alkaloids may interact with cell membranes to alter membrane structure and function, an alteration in the passive permeability of the cell membrane to methotrexate is excluded as an important factor in these studies since a change in the passive diffusion of methotrexatc should be accompanied by an alteration in bidirectional fluxes but the unidirectional influx of methotrexate is unchanged under conditions in which the unidirectional efflux process is markedly retarded. It is unlikely that the vincristine effect on methotrexate uptake is secondary to the metabolic consequences of the mitotic block (34-36), since the interval between exposure of cells to vincristine and st,imulation of net methotrexate uptake is brief (4 min.). The nature of the vincristine-induced alteration in the uptake of methotrexate may contribute to a further understanding of the mechanism(s) by which methotresate inhibits DNA synthesis. Studies from this laboratory suggest that inhibition of deoxyuridine incorporation into DNA by methotrexate is influenced by the level of intracellular mcthotrexate in excess of that required for stoichiometric binding to dihydrofolate reductase.
When methotrexate uptake is enhanced by vincristine in Ehrlich ascites tumor cells, inhibition of DNA synthesis is increased, while vincristine alone has no effect under similar conditions (7). This may account for the increased survival of mice bearing t'he L1210 ascitic tumor when vincristine is administcrcd with methotresate (6) while vincristine alone has little effect (6,27).
This interaction between methotrexatc and the vinca alkaloids should provide a rational basis for the design of synergistic chemotherapeutic regimens with these agents. Conversely, vincristine might antagonize the effectiveness of other cytotoxic agents for which the maintenance of a high intracellular drug level is dependent upon cellular energy metabolism.
Further, the apparent difference between the effects of vincristine on net uptake of methotrexate in L-cells as compared to the tumor cells studied suggests specificity among different tissues and raises the possibility that vincristine may increase the cytotoxicity of mcthotrexate to the tumor without a comparable increase in toxicity to susceptible host tissues.
ncl;nowlecZgment-The excellent technical assistance of Mrs. Sharon Loftfield is acknowledged.